Various embodiments for reducing resource allocations and application instances in diagonal scaling in a distributed computing environment. Resource consumption of each one of a plurality of application instances is tracked, and the tracked resource consumption is compared to a resource allocation of each one of the plurality of application instances. A plurality of resource reduction operations for allocation of resources assigned to each one of the plurality of application instances is computed and applied. Idle application instances of the plurality of application instances are identified, and one or more of the identified idle application instances are terminated, thereby optimizing application efficiency and resource utilization in the distributed computing environment.